{"id":"7fd655a8-835a-4065-83fb-ede3cd5c1c85","arxiv_id":"2507.04757","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A census of Cepheids with periods above 50 days in the Milky Way, Magellanic Clouds, M31, and M33 finds only one clear infrared excess, in the LMC.","lead":"Astronomers fitted the light output of 55 very long-period Cepheid stars in five galaxies to stellar-atmosphere models, searching for extra infrared light from dust. Almost none show an infrared excess, which matters for how these pulsating stars are used to measure cosmic distances and the Hubble constant.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The null result for IR excess in external galaxies is not yet robust: MIR coverage is incomplete and no detection limits are quantified, so the claimed contrast with the Milky Way may reflect selection effects rather than astrophysics.","rationale":"The paper's main contribution is a careful SED-fitting census of long-period Cepheids, and the headline claim is explicitly a null result: no significant IR excess in the SMC, M31, and M33, with only one detection in the LMC. The reader's weakest_assumption correctly identified that the null result depends on the completeness and sensitivity of the available MIR photometry and that the paper itself acknowledges missing NIR/MIR data for four objects. My stress-test agrees with that assessment and sharpens it: the absence of quantified detection limits is the key load-bearing gap. Without an injection-recovery test, one cannot distinguish 'no dust' from 'dust too faint to detect at these distances.' The paper's own admission in Section 4.3 supports this concern, and the large χ²_r values for several external-galaxy fits suggest photometric scatter that could mask small excesses. However, the paper is appropriately cautious in its wording and does provide image inspection as a sanity check. The concern does not invalidate the null result; it limits its strength. The reader's verdict of CONDITIONAL is appropriate, with the condition being that the null result be interpreted as an upper limit pending a sensitivity analysis. Therefore I recommend no change to the reader's verdict, and I agree that the weakest assumption is the completeness of the MIR data and the detectability of an excess if present.","tokens_in":913,"tokens_out":1268,"duration_ms":67246,"concrete_test":"Use the public SED fits and photometry (Zenodo 15422721) to run an injection-recovery experiment. For every star without a claimed IR excess, add a synthetic dust component with the same fractional excess as LMC-CEP-0619, scaled to that star's luminosity and distance, generate synthetic photometry at exactly the wavelengths and uncertainties of the real data, and rerun the MoD fitting with the same BIC criterion. Report the recovery fraction separately for SMC, LMC, M31, and M33. If the recovery fraction is below 80% in M31 or M33, or if any of the four stars without MIR photometry are counted as non-detections despite being impossible to test, the null result is not informative at the level of the detected LMC excess, and the abstract should be revised to state an upper limit rather than an absence.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that only one long-period Cepheid in the LMC (and none in the SMC, M31, or M33) shows a significant infrared excess, implying that IR excess is not a common phenomenon in these systems and is unlikely to bias the Cepheid distance scale. This is a null result, and its weight depends entirely on the sensitivity of the available MIR data to the kind of excess detected in the Milky Way. That sensitivity is never quantified. Section 4.3 notes that four objects have no photometry beyond the NIR, and the observation is correctly qualified with 'absence of proof is not the proof of absence,' but the same concern applies more broadly: for the remaining M31 and M33 stars, WISE W3/W4 and IRAC measurements are often shallow, single-epoch, or absent, and the SED fits have large reduced chi-squared values (Table A.2), indicating that individual photometric points carry large uncertainties. A circumstellar excess similar to that of LMC-CEP-0619 (T_d ~ 925 K, τ_V ~ 0.08; Table 2) would contribute only a small fraction of the stellar flux at 8-24 μm; whether such an excess would be statistically preferred by the BIC given the actual error bars is never tested. Furthermore, the contrast with the Milky Way is partly a comparison with the shorter-period Galactic sample from G20: for P>50 d, the MW rate in Table 4 is only 0/5-1/8, which is not strikingly higher than the external-galaxy rates. Thus the abstract's phrasing overstates the difference, and the data as presented cannot discriminate between a true absence of dust in external galaxies and a detectability limit imposed by distance and photometric depth.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs and fits SEDs for 55 long-period (P > 50 d) classical Cepheid candidates in the Milky Way, LMC, SMC, M31, and M33, using literature photometry and MARCS model atmospheres, optionally adding a dust shell. Distances and reddenings are adopted from the literature, and WISE/IRAC images are inspected to verify candidate infrared excesses. The principal result is that only LMC-CEP-0619 shows a confirmed IR excess, with II Car in the MW being an additional but possibly Type II Cepheid case; no IR excess is found in the SMC, M31, or M33. The paper also places the stars in the HRD and derives PL and PR classifications for the MW/Bulge objects, concluding that the impact of IR excess on the Cepheid distance scale is likely small.","tokens_in":23803,"tokens_out":8665,"duration_ms":100007,"significance":"If the null result is robust, the paper provides an important constraint: warm circumstellar dust around long-period Cepheids is not common in external galaxies, and the Cepheid distance scale is unlikely to be significantly biased by IR excess. The analysis is careful and homogeneous across five galaxies, and the two confirmed excesses (LMC-CEP-0619 and II Car) are supported by independent WISE/IRAC image inspection. The public release of the SED fits and the use of established fitting tools are strengths. However, the weight of the central null claim depends on the sensitivity of the available MIR data, which is never quantified; the paper itself acknowledges the relevant caveat in Sec. 4.3. As it stands, the abstract overstates the contrast with the Milky Way for the same period range.","major_comments":[{"comment":"The central null claim is not accompanied by a quantified detection threshold. The paper states in Sec. 4.3 that four objects have no photometry beyond the NIR and that 'the absence of proof for infrared excess is not the proof of absence,' but this caveat applies more broadly: for the remaining M31 and M33 stars the MIR data are often single-epoch and shallow, and the SED fits show large reduced chi-squared values (Table A.2). No test is made of whether an excess like that of LMC-CEP-0619 (T_d ~ 925 K, tau_V ~ 0.08; Table 2) would be statistically preferred given the actual photometric error bars. I request an injection/recovery test or, at minimum, upper limits on tau_V for each non-detection, and a corresponding softening of the abstract and conclusion.","section":"Sec. 4.3, Table 4"},{"comment":"The claimed contrast with the Milky Way is overstated. For P > 50 d, Table 4 gives MW 0/5-1/8, LMC 1/9, SMC 0/7, and M31+M33 0/19; these rates are statistically indistinguishable from one another. The 'contrary to earlier work' phrasing in the abstract implicitly compares with the shorter-period Galactic sample (16/350 for P < 50 d) rather than with the long-period sample studied here. The abstract and summary should be rephrased to report that no strong excess is seen in the long-period external-galaxy sample, and that a comparison with MW long-period Cepheids is not yet statistically meaningful.","section":"Abstract, Table 4"},{"comment":"The Introduction calls this 'a complete sample of long-period Cepheids,' but Sec. 2 states that known CCs in M31 and M33 with P > 50 d are not included if they are absent from the Gaia vari_cepheid table. The sample is therefore not a complete census of long-period Cepheids in those galaxies, and the null rates in Table 4 are rates for a Gaia-selected subsample. The completeness statement should be corrected and the selection function should be stated explicitly as a limitation on the external-galaxy null result.","section":"Sec. 2 / Sec. 1"},{"comment":"The BIC comparison is not made on equal footing. The 'no dust' model in Table A.2 excludes clipped outliers, while the dust model appears to use a different treatment of the photometric points; for LMC-CEP-0619 the reduced chi-squared is actually larger for the dust model (54.8) than for the no-dust model (51.3), yet the dust model is preferred by BIC. Because six of the eight BIC-preferred candidates are later rejected on the basis of image inspection, the statistical selection step is not the decisive evidence. I recommend reporting BIC or likelihood values computed with identical outlier handling and identical data sets, and specifying the criterion used to call an image detection 'associated' with the central star.","section":"Sec. 4.3, Table 2"},{"comment":"Distance errors are deliberately not propagated into the quoted luminosity errors (Table A.2 note). For MW stars with fractional distance uncertainties of 10-30% (e.g., ATO093m31 at 15.78 +/- 2.58 kpc), this affects the HRD, PL, and PR classifications in Table 3, and hence the denominator of the MW >50 d excess rate in Table 4. The alternative-distance runs in Sec. 4.4 address this partially, but the standard-model classifications in Table 3 are used without this uncertainty, so the MW comparison rate should carry a corresponding caveat.","section":"Table A.2, Sec. 4.4"}],"minor_comments":[{"comment":"The description of how Teff errors are derived ('models within a certain range above the minimum') should specify the adopted Delta-chi-squared threshold and how the 125 K interpolation step enters the quoted uncertainties.","section":"Sec. 3.3"},{"comment":"The W3/W4 quality cuts are given in magnitude-error terms, but the number of epochs and the pulsation phase of single-epoch MIR measurements are not documented; phase-dependent photometry could masquerade as scatter or as a false excess.","section":"Table B.1"},{"comment":"The visual classification of 'not clearly associated' would be more reproducible with a quantitative measure, such as PSF-matched aperture photometry at the expected stellar position versus the local background.","section":"Figs. C.2 and C.3"},{"comment":"The phrase 'very long-period' is not defined; since the sample starts at 50 d and includes many objects below the usual 80 d ultra-long-period threshold, 'long-period' may be more accurate.","section":"Title / Sec. 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the journal's scope and the analysis is generally careful, with useful imaging verification of the two confirmed excesses. The main issue is not circularity or technical execution but the strength of the null claim: the lack of a quantified detection sensitivity and the incomplete M31/M33 sample make the abstract's wording too strong. I would be satisfied after the authors add detection-limit estimates or clearly bounded upper limits, correct the completeness wording, and revise the MW comparison statement. These are fixable within the manuscript's scope, so I do not recommend rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read it for the census, not the cosmology. The paper extends Groenewegen's SED-fitting program to 55 long-period Cepheids across the MW, LMC, SMC, M31, and M33, and finds only one convincing IR excess in the LMC (plus II Car, likely a T2C) after imaging checks. That is the new result, and it is genuinely useful: it weakens the idea that dust or free-free emission biases Cepheid PL relations in a metallicity-dependent way.\n\nThe strengths are real. The photometry is assembled carefully from the literature, the fits use a consistent model grid, and the BIC comparison between dusty and dust-free models is transparent. The WISE/IRAC image inspection is the most valuable step because it separates true circumstellar excess from diffuse background emission; for several M31 and M33 candidates the imaging correctly shows the excess is not associated with the star.\n\nThe stress-test note is partly right. The null result is not as robust as the abstract implies. Four objects have no photometry beyond the NIR, and more generally the M31/M33 MIR data are often shallow or single-epoch. The paper never quantifies the minimum dust optical depth that would have been detectable given the actual error bars, so the M31/M33 zero detections are upper limits, not firm zeros. The BIC is applied 'with flexibility', fine for discovery but not for statistical significance. And the MW contrast in Table 4 is not dramatic: for P>50 d the MW rate is only 0/5-1/8, so the abstract's phrasing overstates the difference.\n\nBut the stress-test goes too far when it says the data cannot discriminate at all. The fits and imaging do rule out the kind of strong excess seen in LMC-CEP-0619; the issue is only for weak excess. Distance errors not being propagated into luminosity errors is a minor caveat and does not affect the excess claim.\n\nThe paper deserves a serious referee. I would ask the author to quantify detection limits, soften the abstract, and explicitly flag the four NIR-only objects in the summary table. I would cite Table 4 as the cleanest statement of the result, with the caveat attached.","headline":"A careful multi-galaxy census that confirms IR excess is rare in external-gala xy Cepheids, though the null result is less airtight than the abstract suggests.","tokens_in":24347,"tokens_out":1942,"would_cite":true,"duration_ms":22358,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["97.30.Gj"],"model":"deepseek-v4-flash","headline":"Among 35 very long-period Cepheids in the Magellanic Clouds, M31, and M33, only one shows a significant infrared excess — dust is unlikely to bias the Cepheid distance scale.","keywords":["spectral energy distributions","classical Cepheids","infrared excess","circumstellar dust","Magellanic Clouds","M31","M33","distance scale"],"falsifier":"Point new mid- or far-infrared observations at the 19 Cepheids in M31 and M33 — most directly the four that currently lack data beyond the near-infrared (M31-PSO009.76, M33-013331, M33-V00021, M33-013405) — for example with JWST/MIRI imaging or spectroscopy; if several of these stars show excess emission above the best-fitting photosphere, or silicate features appear in the spectra, the conclusion that long-period Cepheids in external galaxies are dust-free would be overturned.","tokens_in":23211,"feed_emoji":"🌟","tokens_out":12490,"duration_ms":113919,"temperature":0.7,"pith_summary":"The paper asks whether the very longest-period classical Cepheids — stars with pulsation periods above 50 days that are used as standard candles — are commonly surrounded by dust that adds extra infrared light to their spectra. It builds spectral energy distributions (their light spread across wavelengths) for 55 such stars in the Milky Way, the Large and Small Magellanic Clouds, M31, and M33, then fits each with model stellar atmospheres, adding a dust shell only when the data require it. Across the 35 stars in the external galaxies, exactly one clear infrared excess survives (LMC-CEP-0619), with none in the SMC, M31, or M33, and one further excess in the Milky Way (II Car) whose status as a classical Cepheid is unclear. A sympathetic reader would care because infrared excess can contaminate the near- and mid-infrared photometry that feeds period-luminosity relations; the paper concludes that for these long-period Cepheids the impact on the distance scale should be low at best.","feed_headline":"Just one of 35 long-period Cepheids shows an infrared excess","feed_subtitle":"With dust absent around these standard candles, the extragalactic distance scale avoids a potential bias.","key_machinery":"The central tool is the spectral energy distribution assembled from literature photometry spanning ultraviolet to mid-infrared wavelengths (GALEX, Gaia, 2MASS, VMC, WISE, IRAC, MIPS, and others). Each SED is fitted with MARCS model atmospheres at a fixed adopted distance and reddening, yielding the best luminosity and effective temperature; for stars where long-wavelength flux remains unexplained, a spherical dust shell is added with the dust optical depth and inner-edge temperature as free parameters, and the Bayesian information criterion decides whether the dust model is genuinely better. The decisive step is image inspection: WISE W1/W3 and IRAC channels 1 and 4 cut-outs are used to verify that claimed excess emission actually coincides with the star, which removes the M31 and M33 candidates whose infrared flux turned out to be diffuse or blended background instead.","core_discovery":"On its own terms, the paper establishes that very long-period classical Cepheids in external galaxies do not commonly present an infrared excess in their spectral energy distributions. Fitting MARCS model atmospheres with the MoD radiative-transfer code to photometry from the ultraviolet to 24 micrometers, and checking candidate excesses against WISE and IRAC images to discard blended or diffuse background emission, the author finds one significant excess in the LMC (LMC-CEP-0619) and none among the seven SMC, twelve M31, and seven M33 objects; in the Milky Way, II Car shows an excess but it is unclear whether it is a classical Cepheid or a Type II Cepheid. This runs contrary to earlier Galactic work that hinted infrared excess might be more prominent in Milky Way Cepheids than in the Magellanic Clouds. The same fits locate nearly all objects inside the classical Cepheid instability strip in the Hertzsprung-Russell diagram, with masses near 10-15 solar masses, while the outliers are largely the objects already re-classified as Type II Cepheids; for Milky Way objects, changing the adopted distance or reddening can move stars in and out of the strip, so those positions carry larger uncertainty.","pith_inferences":["A direct test the paper leaves implicit: longer-wavelength photometry for the four M31 and M33 objects that currently stop at the near-infrared could still reveal warm dust, so the zero-detection count in those galaxies is an upper limit until such data exist.","If the genuine excesses (LMC-CEP-0619 and II Car) turn out to come from free-free emission of ionized gas rather than dust — an option the paper itself discusses — then the interesting comparison across galaxies is not dust formation but the prevalence of circumstellar ionized gas around the most luminous Cepheids.","Extending the same fitting pipeline to the shorter-period Cepheids in M31 and M33 (left unstudied here) would test whether the roughly 5 percent excess fraction seen in Milky Way stars is universal; a null there would make the Milky Way the anomaly rather than the rule."],"forward_implications":["If the near-null holds, mid-infrared photometry of long-period Cepheids in the LMC, SMC, M31, and M33 is not contaminated by circumstellar dust, so period-luminosity relations built from those bands carry no dust-excess bias for these objects.","The earlier hint that infrared excess is more prominent among Milky Way Cepheids than in the Magellanic Clouds is not confirmed for the longest-period stars, weakening the case that metallicity drives the fraction of dusty Cepheids.","The HRD placement inside the classical Cepheid instability strip for most objects supports treating the >50-day variables as classical Cepheids with masses near 10-15 solar masses, while the confirmed outliers clarify which catalog entries are actually Type II Cepheids.","For the Milky Way subsample, distance and reddening uncertainties move stars substantially in the HRD, so improved parallaxes (such as Gaia DR4) are needed before those objects can pin down luminosities or effective temperatures tightly."],"supporting_citations":[{"why":"The earlier SED study of 477 Galactic Cepheids that reported infrared excess in a subset; the Milky Way baseline this paper tests against.","marker":"Groenewegen (2020a)"},{"why":"The preceding SED analysis of SMC and LMC Cepheids that found just one infrared excess, which this paper extends to M31 and M33.","marker":"Groenewegen & Lub (2023)"},{"why":"Supplies the long-period Galactic Cepheids from the OGLE all-Galactic catalog of 3666 objects.","marker":"Pietrukowicz et al. (2021)"},{"why":"The OGLE-IV catalog that provides the LMC and SMC Cepheids with periods above 50 days.","marker":"Soszyński et al. (2019)"},{"why":"The Gaia DR3 vari_cepheid table that defines the unified 55-object sample and the re-classification of several stars as Type II Cepheids.","marker":"Ripepi et al. (2023)"},{"why":"Eclipsing-binary distance to the LMC (49.59 kpc) used to convert fitted fluxes to luminosities.","marker":"Pietrzyński et al. (2019)"},{"why":"Eclipsing-binary distance to the SMC (62.44 kpc) used to convert fitted fluxes to luminosities.","marker":"Graczyk et al. (2020)"},{"why":"The MARCS model atmosphere grid that supplies the stellar photosphere spectra for the fits.","marker":"Gustafsson et al. (2008)"},{"why":"The DUSTY radiative-transfer code inside the MoD fitting routine used to add and test dust shells.","marker":"Ivezić et al. (1999)"},{"why":"The theoretical instability-strip edge positions against which the HRD locations are judged.","marker":"De Somma et al. (2021)"}],"fun_headline_variants":["Rare infrared excess in long-period Cepheids","IR excess nearly absent in long-period Cepheids","One IR excess found across 35 long-period Cepheids","Long-period Cepheids show few infrared excesses","Cepheid IR excess: a lonely LMC star stands out"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The finding that no infrared excess exists in most of these stars rests on the available mid-infrared photometry being sensitive enough to reveal one — the paper itself states that the absence of proof of infrared excess is not proof of absence, and four of the M31 and M33 objects have no photometry beyond the near-infrared.","fun_headline_variants_meta":{"raw":{"variants":["Rare infrared excess in long-period Cepheids","IR excess nearly absent in long-period Cepheids","One IR excess found across 35 long-period Cepheids","Long-period Cepheids show few infrared excesses","Cepheid IR excess: a lonely LMC star stands out"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000166,"raw_usage":{"total_tokens":1336,"prompt_tokens":1108,"completion_tokens":228,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":724,"completion_tokens_details":{"reasoning_tokens":147}},"tokens_in":724,"tokens_out":228,"duration_ms":3407,"temperature":1.0,"reasoning_tokens":147,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:39:57.254175+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Point new mid- or far-infrared observations at the 19 Cepheids in M31 and M33 — most directly the four that currently lack data beyond the near-infrared (M31-PSO009.76, M33-013331, M33-V00021, M33-013405) — for example with JWST/MIRI imaging or spectroscopy; if several of these stars show excess emission above the best-fitting photosphere, or silicate features appear in the spectra, the conclusion that long-period Cepheids in external galaxies are dust-free would be overturned.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The preceding SED analysis of SMC and LMC Cepheids that found just one infrared excess, which this paper extends to M31 and M33."},{"cited_title":"2021, , 71, 205","cited_arxiv_id":null,"evidence_quote":"Supplies the long-period Galactic Cepheids from the OGLE all-Galactic catalog of 3666 objects."},{"cited_title":"2023, , 674, A17","cited_arxiv_id":null,"evidence_quote":"The Gaia DR3 vari_cepheid table that defines the unified 55-object sample and the re-classification of several stars as Type II Cepheids."}],"review_version":1}